Creating More Efficient Approaches to Medical Research Using Real-World Data

Medical researchers face a collective problem of too many questions, too little time. “As someone who practices medicine,” says Associate Professor of Neurology Brian Mac Grory, “I’ve noticed that there’s a huge amount of questions that need to be answered. We need to figure out how to treat people with medicines, how to treat people with surgeries and how to understand what causes different diseases. We can’t possibly answer all of them by designing new studies and clinical trials every time.” 

Every time researchers do conduct a study, it generates new questions that need to be answered. Thus, Mac Grory says, “we’re playing this continuous game of catch-up in medical science.”

Image

Could there be a more efficient approach to answering medical questions that can inform physicians, policymakers, business and the public? Leveraging existing data is a promising way to get answers without requiring a time-consuming and expensive clinical trial. 

Mac Grory explains that there are two broad types of research that doctors do. “One is research where we identify a question. Either we want to know more about a disease or we want to test a new treatment, and we design a study from the very start. We plan everything out in advance, identify patients and collect data explicitly for that study, and we use the data we collect to answer questions.”

The second type of research involves using data that’s been collected for other reasons. “All of us in our day-by-day lives have many thousands of pieces of data collected about us,” Mac Grory points out. “Data that’s collected as part of medical care, for example, [comes from] billing, pharmacy, quality assurance, documentation to help coordinate between departments of the hospital … all these data are being collected for other reasons, but medical research can be done using these data. We can potentially do a study in a very efficient manner.”

If researchers want to study a particular rare condition, they can spend tens of millions of dollars and five years identifying a group of people with that condition, or maybe they can access large databases that have already collected data on people with that condition and use the data for medical research — “provided all the right safeguards are there, provided you’re respecting privacy and anonymity and any regulations and so forth,” Mac Grory emphasizes.

For the past few years, Mac Grory and some colleagues across Duke have been working on using existing data to answer questions “in the best way possible,” he says. Last year, they tapped into a funding opportunity from the Office of the Provost and were selected for a Multiyear Interdisciplinary Hub grant for research, education and engagement. They called their group ORBIT, which stands for Observational Research Building Interdisciplinary Therapeutic Advances.

ORBIT’s two main objectives are to develop techniques and tools to use real-world data to answer questions that would otherwise require a randomized controlled trial, and to make such trials more efficient by using real-world data overlaid with techniques in AI and data science.

Image

“The people that I’m working with are Laine Thomas, who’s a professor of biostatistics, Fan Li, a professor of statistical science, Emily O’Brien, a population health scientist, Ryan McDevitt, an economist, Anqi Zhao, a decision scientist, Ricardo Henao, who’s a statistician and a computer scientist, and Jay Lusk, a health services researcher and preventive medicine physician,” says Mac Grory. Their collaboration “evolved organically over time based on conversations we were having about the unmet need for real-world evidence generation.”

Since the grant began last summer, the group has hired four statisticians and fleshed out its entire team. “One of the innovative things about ORBIT is that we have economists as a central part of the team,” Mac Grory says. “This elevates our work by bringing in new voices and perspectives to bear on difficult problems in medicine.”

They’ve submitted eight grant proposals (and already received two awards), published three papers, organized a Winter School in Real-World Data Analytics and hosted monthly seminars.

“We’ve done lots of different activities for research, education and outreach across the Duke community,” Mac Grory notes. One of these activities is launching a pilot grant program, which offers one-year awards of $10,000 to junior investigators at Duke. “People can do anything that involves real-world evidence generation using real-world data,” he says. “We’re also going to provide mentorship and other resources for people who are selected for the grants.” He hopes the program will help ORBIT identify people across the university who’d like to collaborate in other ways.

Image

In addition to the provost’s grant, the Duke Clinical Research Institute has been a vital partner to the ORBIT Hub. “One of the many things that they’ve done is given us a project manager for our research program, Marcus Layer, who is an exceptional leader and has really transformed our activities,” Mac Grory shares. “Now we have someone who brings us all together and organizes and leads everything.”

Through the Office for Research and Innovation, project manager Lauren Whittaker is leading the hub’s educational programs. “She’s really doing a fantastic job,” says Mac Grory. “Before the Hubs grant, we were working in a disorganized fashion, and we're now all working together to achieve our objectives.”

About the Interdisciplinary Hubs

Last year, the Office of the Provost selected three faculty-led projects to receive Multiyear Interdisciplinary Hubs grants for research, education and engagement. The grants are intended to galvanize emerging intellectual connections, research collaborations and links to external partners, as well as create compelling pathways of inquiry for students at all levels. 

“These Interdisciplinary Hubs are an important new mechanism for enabling exceptionally promising projects to take the next leap forward and a key strategic investment at an uncertain time for research funding,” said Provost Alec D. Gallimore.

Besides ORBIT, the other two selected projects are the Duke Critical Minerals Hub and the Interdisciplinary Hub for Rural Health Equity. Each hub receives three years of support along with project management effort from the Office of Research & Innovation’s Research Project Management Core.

Main image: In the new ORBIT Hub, investigators with complementary expertise are working together to overcome pitfalls with randomized controlled trials (RCT).